The present inventions are related to systems and methods for preparing dibit signals that may be used for estimating non-linear parameters, and more particularly, to systems and methods for correcting a dibit signal to remove DC offset and/or restore the baseline.
Dibit signals are used in a variety of processing scenarios. For example, a dibit signal may be used to estimate a variety of non-linear parameters in a hard disk drive system utilizing longitudinal or perpendicular recording approaches. Such non-linear parameters include non-linear transition shifts, magneto-resistive asymmetry, channel linear density, overwrite, and the like. These parameters help to understand the types and strengths of nonlinear distortions present in the recording channel and also for developing approaches for mitigating these distortions. Thus, dibit extraction circuits may be included in a variety of circuits. However, it is often difficult to use such extracted dibit signals due to DC offset or distorted baseline exhibited in the dibit signal. This DC offset is caused by the non-zero cross-correlation property of the pseudo-random bit sequence (PRBS) used for extracting the dibit in conjunction with the non-zero DC property of the recording channel. This DC offset issue is more serious in perpendicular recording, as compared to longitudinal recording. FIG. 1 shows an exemplary prior art dibit signal 100. Dibit signal 100 includes a main lobe 110 with a number of echoes 130 disposed on either side of main lobe 110. As shown, main lobe 110 exhibits a magnitude significantly larger than any of echoes 130. As shown, a DC offset 120 may be so great that dibit signal 100 may be unusable.
Some approaches rely on a high pass filter present in the analog front end (i.e., AC-coupling) to remove the DC offset. However, in many cases, the pole of the high pass filter is very narrow and focused on removing only the DC frequency data. Because the pole is so narrow, there is often a transient affect that results in an undesired modification to the dibit signal while not solving the DC offset problem. Where the pole is made broader, too much information is removed from the dibit signal. Other approaches use a non-echo region (i.e., a flat region of the dibit signal) to estimate the DC offset. This estimated DC offset is then added/subtracted from the entire dibit signal. This approach does not work well, however, because the non-echo region is generally not perfectly flat due to AC coupling and presence of echoes. Other approaches perform dibit correction where the post process bits are represented by the following equations:
                    h        b            ⁡              [        k        ]              =                  (                              p            ⁡                          [              k              ]                                -                                    1              N                        ⁢                                          ∑                l                            ⁢                              p                ⁡                                  [                  l                  ]                                                                    )            ⁢              N                  N          +          1                      ,for DC-free channel; and
                    h        b            ⁡              [        k        ]              =                  (                              p            ⁡                          [              k              ]                                +                                    ∑              l                        ⁢                          p              ⁡                              [                l                ]                                                    )            ⁢              N                  N          +          1                      ,for channels with DC.Such approaches provide accurate correction of the dibit signal only where no non-linear distortions or distortion from AC coupling are present.
Hence, for at least the aforementioned reasons, there exists a need in the art for advanced systems and methods for correcting dibit signals.